1. The Field of the Invention
The present invention relates to an information retrieval system and in particular relates to an information retrieval system to be used for networked information processing system.
Recently, it is desired in large firms to form therein networked and hierarchically constructed information processing systems. Such information processing systems enhance the information-orientation of the firms. Further, an information retrieval system is also desired, which system has flexibility such that an information retrieval for the above type of information systems may be easily and efficiently performed.
2. The Related Art
A networked information processing system has the following construction, for example: There is an general-information server apparatus which manages general information concerning the entirety of a firm. There are further sectional-information server apparatuses, which manage sectional information used in each section, that is each part of the firm. This sectional information comprises a partial set of the above general information, for example. There are further personal-information server apparatuses, which manage personal information which each employee has, the employee belonging to a section of the firm. This personal information comprises a partial set of either the above general information or the above sectional information, for example. Then, these general-information server apparatus, sectional-information server apparatuses and personal-information server apparatuses are interconnected through a line network such as a Local Area Network (LAN, hereinafter).
An information retrieval system in the related art for such a networked information processing system will now be described. In the retrieval system, an information unit which a user desires to obtain is directly accessed.
Such a retrieval system may be implemented by a command provided in an operating system employed in an on-the-market personal computer or workstation, for example. Further, such a retrieval system may comprise a utility function of an on-the-market full-screen editor. However, since the retrieval speed in such retrieval systems is not fast, a search system such as that referred to as a keyword search or that referred to as a full-text search is devised so that improvement of the retrieval speed is obtained.
In order to implement the keyword search, keywords are automatically extracted from information units in accordance with a keyword dictionary. Each of the information units is made by converting an individual document into the corresponding electronic form. The keyword dictionary comprises a set of keywords which is to be used for search. Then, an inverted file is automatically produced comprising the thus extracted keywords and record numbers. Each of the record numbers indicates a position, on an information storing medium, at which position an information unit is stored., the information unit corresponding to the extracted keyword. Then, at the time of retrieval, a keyword is specified by a user and then the retrieval system searches the keywords in the inverted file for the specified keyword.
Other terms such as "index file" may be used instead of "inverted file". In such a keyword search, it is not possible to retrieve an information unit using a keyword which is not included in the keyword dictionary (free keyword). Further, it is necessary to regenerate the inverted file(s) if the information unit and/or the keyword dictionary is revised.
The technology and form of above-mentioned full-text search may differ slightly when applied to different information processing systems. In order to implement the full-text search, a retrieval file is produced by compressing the information units, for example. Even during this, compressing, almost all of the original contents of the information units are maintained. Other terms besides "retrieval file" may be used to denote the same entity in different information retrieval systems.
Advantages obtained from the full-text search will now be described. It solves the following problems occurring in the keyword search: It is not possible to search the entirety of the information units; and it is necessary to regenerate the inverted file if the information units are updated as mentioned above. Further, the above-mentioned full-text search using the retrieval file, solves the problems in a primitive full-text search not using a retrieval file wherein considerable time is required for the primitive full-text search.